Available online at www.sciencedirect.com

ng xx (2012) xxx–xxx

Magnetic Resonance Imagi

An intravascular loopless monopole antenna for vessel wallMR imaging at 3.0 T☆

Hongyang Yuan a, Xing Lv a, Xiaohai Ma b, Rui Zhang a, Youyi Fu c, Xuedong Yang d,Xiaoying Wang c, d, Zhaoqi Zhang b, Jue Zhang a, c,⁎, Jing Fang a, c

aCollege of Engineering, Peking University, Beijing 100871, ChinabDepartment of Radiology, Beijing Anzhen Hospital, Beijing 100029, China

cAcademy of Advanced Interdisciplinary Studies, Peking University, Beijing 100871, ChinadDepartment of Radiology, Peking University First Hospital, Beijing 100034, China

Received 31 March 2012; revised 5 June 2012; accepted 26 June 2012

Abstract

The purpose of this study was to develop a novel intravascular loopless monopole antenna (ILMA) design specifically for imaging ofsmall vessel walls. The ILMA consisted of an unshielded, low-friction guide wire and a tuning/matching box. The material of the guide wirewas nitinol and it was coated with polyurethane. Because the guide wire was unshielded, it could be made thinner than the coaxial cable-based loopless intravascular antenna design. The material of the box was aluminum. In this study, the diameter of the guide wire was 0.5 mmand the length was 58.7 mm. The ILMA was used as a receiving antenna and body coil for transmission. To verify the feasibility of theILMA, in vitro and in vivo experiments were performed on a 3.0-T magnetic resonance (MR) scanner. In vitro tests using the ILMAindicated that the proposed design could be used to image target vessel walls with a spatial resolution of 313 μm at the frequency codingdirection and more than 100 mm of longitudinal coverage. In vivo tests demonstrated that the images showed the vessel walls clearly byusing the ILMA and also indicated that the ILMA could be used for small vessels. The proposed antenna may therefore be utilized to promoteMR-based diagnoses and therapeutic solutions for cardiovascular atherosclerotic diseases.© 2012 Elsevier Inc. All rights reserved.

Keywords: MRI; Intravascular antenna; Vessel wall; ILMA

1. Introduction

Atherosclerotic cardiovascular disease remains the lead-ing cause of death in the world today [1]. Traditionally,X-ray technology has been used to diagnose and treat thisdisease. However, this method provides minimal informa-tion regarding vessel walls [2]. New techniques utilizingintravascular ultrasound technology can provide information

☆ This research was supported by the National Natural Science Fund ofChina (FS30870733).

⁎ Corresponding author. College of Engineering, Peking UniversityBeijing, 100871, P.R. China. Tel.: +86 1062755036; fax: +86 1062753562

E-mail addresses: yhy_1984@163.com (H. Yuan),lvxingvir@gmail.com (X. Lv), maxi8238@gmail.com (X. Ma),zhangruigt@126.com (R. Zhang), fuyouyi0123@gmail.com (Y. Fu),yangxuedong1@163.com (X. Yang), cjr.wangxiaoying@vip.163.com(X. Wang), zhaoqi5000@vip.sohu.com (Z. Zhang), zhangjue@pku.edu.cn(J. Zhang), jfang@pku.edu.cn (J. Fang).

0730-725X/$ – see front matter © 2012 Elsevier Inc. All rights reserved.doi:10.1016/j.mri.2012.06.032

,.

regarding the vessel walls [3], but this method is oftenlimited by the sound-reflecting calcifications often associat-ed with plaques [4]. Magnetic resonance imaging (MRI) hasshown great promise in vessel wall imaging and has shownpotential to characterize both artery atherosclerosis andaortic stenosis disease [5,6]. At present, however, MRIutilizing surface radio frequency (RF) coils is only suitablefor imaging of superficially located arteries, such as thecarotid [6].

In order to obtain pathological features of deep vesselwalls, such as those of the coronary arteries, the intravascularMR probe, which utilizes a special coil/antenna placed insideor closed to the target artery, has been proposed in previousstudies [7,8]. Intravascular MR probes utilize one of twotypes of receivers: loop coil and loopless antenna [4,9]. Asingle-loop, double-turn coil mounted on a catheter was firstproposed to increase the signal-to-noise ratio (SNR) in 31Pspectra of canine hearts [10]. Subsequently, several types of

Fig. 1. Photo of the ILMA (A) and equivalent circuit diagram for tuning and matching (B). C1, C2, C3 and C4 are used for tuning and matching. D is a PIN diodefor decoupling using bias during the experiments.

2 H. Yuan et al. / Magnetic Resonance Imaging xx (2012) xxx–xxx

small loop coils, commonly embedded inside the catheter,have been utilized for imaging the arterial wall with higherlocal SNR [11–14]. In order to improve the longitudinalcoverage and allow multislice scans, long-loop catheter coildesigns [15] have been used in which the tuning andmatching circuits are placed near the catheter coil with thesensitivity falling off as 1/r2 [16]. Moreover, SNR andflexibility have been further improved by integrating coilsinto inflatable balloon structures [11]. Unfortunately,however, current loop coils are often too rigid and large toinsert into small diseased vessels. In addition, signalsensitivity diminishes very rapidly along the radial directionof the coil [16].

Intravascular loopless antennae utilize much thinner andmore flexible coaxial cable [8,17]. The inner conductors ofthese antennae are extended by about a quarter wavelength toform a “whip” suitable for imaging small vessels [16]. Therelatively short length of the intravascular MR whip,however, results in relatively small longitudinal coverage.

In this study, we propose a flexible, long and thinintravascular loopless monopole antenna (ILMA) designbased on the monopole antenna theory [18]. The ILMAconsists of a flexible wire without shielded cable. Thisdesign enables the guide wire to be sufficiently thin forapplications in small vessels and improves longitudinalcoverage for multislice scanning. To determine the feasibil-ity and validity of this design, this study describes the (1)fabrication of an ILMA; (2) SNR estimation of an ILMA;and (3) validation of the produced ILMA with in vitro and invivo experiments.

Fig. 2. In vitro experimental setup of the MRI of the “vessel wall” using theILMA. The guide wire (black line) is inserted into the infusion tube.

2. Materials and methods

2.1. Fabrication of the ILMA

The proposed ILMA consisted of a thin, unshielded,flexible, low-friction guide wire and a tuning/matching box(Fig. 1). The guide wire was made of a nitinol tapered corewith a polyurethane coating (TERUMO, Tokyo, Japan). Thediameter was 0.5 mm and the length was 58.7 mm. Thematerial of the tuning/matching box was aluminum. TheILMA was used as a receive-only probe, while the RF pulseswere transmitted from the body coil embedded in the MR

scanner. In the tuning/matching circuit, there were threeadjustable capacitors (i.e., C1, C2 and C3) and one fixedcapacitor (i.e., C4). C1 was used to enlarge the reflection gain(S11); C2 was used to tune the coil approximately to theexpected frequency; the function of C3 was to fine tune thepeak of frequency. The value of C4 was 150 pF, which wasdesigned for matching the circuit to 50 Ω with theabovementioned three capacitors. After tuning and match-ing, the resonance frequency was 127.72 MHz for 3 T ofmagnetic intensity and the reflection gain (S11) was −30 dBat the resonance peak. D was a PIN diode, which was usedfor decoupling between the body coil and the ILMA duringRF transmission.

2.2. In vitro experiment

A test phantom consisting of an infusion tube inside aplastic tank filled with saline solution was designed to mimica rabbit body and its abdominal aorta (Fig. 2). The size of theplastic tank was 183.1×111.2×56.7 mm. The infusion tubewas inserted into the holes which were drilled centrally onboth sides of the tank. The infusion tube, with inner and outerdiameters of 3 and 4 mm, respectively, was similar inthickness to that of the coronary artery [19].

The saline solution in the infusion tube was kept flowingto simulate transfusion. Fluid velocity was estimated by acommon flow meter. The guide wire of the ILMA was

Fig. 3. The feasibility of the ILMA was validated by an infusion tube-flow phantom at 3 T. An axial image (A) and a sagittal image (B) of the tube-flow phantomobtained by the ILMA. A signal void at the center of the tube is caused by the flowing void effect. A SNR map of the ILMA is shown in (C). The red arrowindicates the wall of the infusion tube, and the yellow arrow indicates the position of the guide wire. (For interpretation of the references to color in this figurelegend, the reader is referred to the web version of this article.)

3H. Yuan et al. / Magnetic Resonance Imaging xx (2012) xxx–xxx

inserted into the infusion tube and the phantom was placed inthe MR scanner in order to evaluate image quality.

A regular-shaped cucumber was also used to furtherevaluate the tissue resolution and longitudinal coverage ofthe fabricated ILMA. In this test, the guide wire was adheredto the outside surface of the cucumber with a Scotchtape. The cucumber sample was 30.1 mm in diameter and199.9 mm in length.

The MRI parameters for the phantom (axial) were asfollows: a fast spin-echo (FSE) sequence with TR=2400.0 ms, TE=200.0 ms, FOV=8×8 cm, 192×160 pixels,slice thickness=3.0 mm, spacing=1.0 mm, bandwidth=31.3 kHz and NEX=2.0. The imaging parameters for thephantom (sagittal) were as follows: a FSE sequence with TR=2400.0 ms, TE=200.0 ms, FOV=8 cm×8 cm, 256 pixels×192 pixels, slice thickness=2.0 mm, spacing=1.0 mm,bandwidth=31.3 kHz and NEX=2.0. The imaging parametersfor the cucumber (axial and sagittal) were as follows: a FSEsequence with TR=3000.0 ms, TE=83.0 ms, FOV=20×20 cm, 192×160 pixels, slice thickness=4.0 mm, spacing=1.0 mm, bandwidth=15.6 kHz and NEX=1.0. All the in vitroand in vivo experiments were performed on a GE 3.0-TMRI scanner (GE Healthcare, Milwaukee, WI, USA).

2.3. In vivo experiment

Three healthy New Zealand white rabbits were used toverify the validity of the ILMA.Each rabbit was approximately4.0 kg in weight, with an aorta of approximately 3.0 mm indiameter. The experimental protocol was approved by theBeijing Anzhen Hospital Animal Care and Use Committee.

Through a surgical procedure, the ILMA guide wire wasinserted into the abdominal vein through a catheter while therabbit was under general anesthesia.

An axial image was acquired using a 3.0-T eight-channelTORSO Array Coil (GE Healthcare) to determine theanatomic position of the abdominal aorta. Axial and sagittalimages were then acquired by intravenous MRI of theadjacent abdominal aorta using the ILMA device.

For ILMA-based MRI of the rabbit (axial), the imagingparameters were as follows: a FSE sequence with TR=2400.0 ms, TE=149.0 ms, FOV=12×12 cm, 320×192 pixels,

slice thickness=6.0 mm, spacing=1.0 mm, bandwidth=31.2 kHz and NEX=8.0. The sagittal plane imagingparameters were as follows: a FSE sequence with TR=2400.0 ms, TE=170.0 ms, FOV=20×20 cm, 320×192 pixels,slice thickness=2.0 mm, spacing=1.0 mm, bandwidth=31.2 kHz and NEX=8.0. For the rabbit (axial) with theTORSO coil to receive signal, we performed MRI using thefollowing parameters: a FSE sequence with TR=2400.0 ms,TE=149.0 ms, FOV=18×18 cm, 320×192 pixels, slicethickness=6.0 mm, spacing=1.0 mm, bandwidth=31.2 kHzand NEX=2.0.

2.4. SNR Estimation of images obtained by the ILMA

ILMA imaging quality was evaluated by SNR mapscreated with MATLAB software. The SNR was estimatedpixel by pixel based on the following expression: SNR=S/SD, where S is the signal amplitude of the region ofinterest (ROI) in a given image and SD refers to the standarddeviation of noise in the ROI (near the margin of the images).

Furthermore, the resolution of an image in the x direction(frequency coding direction) was obtained by the followingformula: FOVx/Nx, where FOVx is the length of the FOV inthe x direction, Nx is a component of the matrix (matrix=Nx

pixels×Ny pixels). In this work, we used the phantomparameters to evaluate the resolution.

3. Results

In the in vitro experiments, the average velocity of the salineflowing in the infusion tube was (60.0±0.05)×10−4 m/s. Theinfusion tube image obtained by the ILMA and thecorresponding SNR map are presented in Fig. 3. The highestSNR was 130 (Fig. 3C). The ILMA generated high-resolutionimages (313 μm) at the frequency coding direction to allow theplastic “vessel” wall (low signal intensity) and saline solution(high signal intensity) to be distinguished.

The axial cucumber image produced from the ILMAdevice also indicates the feasibility of this device. The tissueis clearly shown in the image (Fig. 4A). The sagittalcucumber image indicates that the large longitudinalcoverage of the ILMA is more than 100 mm (Fig. 4B).

Fig. 4. An axial image (A) and a sagittal image (B) of the cucumber obtained using the ILMA at 3 T. The sagittal image indicates that the longitudinal coverage ofthe ILMA is more than 100 mm (the length of the cucumber is 199.9 mm). A SNR map of the ILMA is shown in (C). The yellow arrow indicates the position ofthe guide wire. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

4 H. Yuan et al. / Magnetic Resonance Imaging xx (2012) xxx–xxx

In the vivo study, both the axial and sagittal imagesacquired by the ILMA displayed the vessel wall clearly(Fig. 5C and D). A reference image acquired by an eight-channel TORSO Array Coil (Fig. 5A and B) indicated thatthe position of the abdominal aorta obtained by the ILMAwas identical to that of the TORSO Array Coil. Therefore,the images clearly depicting the abdominal aorta wall of therabbit are reliable as determined by the ILMA.

4. Discussion

This study described and validated a clinical-size,unshielded, flexible and low-friction guide wire-basedintravascular loopless monopole antenna for MRI. Theunshielded guide wire of the proposed ILMA enables a thinand flexible antenna to be inserted into small and tortuousvessels. Moreover, the ILMA can provide large longitudinalcoverage for multislice imaging.

To date, coaxial cable-based loopless intravascularantennae utilize a dipole antenna with two radiating elements[6,17]. In contrast, the proposed ILMA is a typical monopoleantenna with only one radiating element with no need of ashielded wire. In this case, the guide wire can be made verythin and may be more applicable for small vessels forpotential clinical use.

Fig. 5. An image of the abdominal aorta of the rabbit by 3.0-T eight-channel TORSthe ILMA (C and D). This result indicates that the position of the aorta as determinedshown in (B). The red arrow indicates the abdominal aorta lumen of the rabbit and ththe abdominal aorta obtained by the ILMA (FSE T2) indicate that the ILMA is feaswhile the yellow arrow indicates the position of the guide wire. The red arrow poiinterpretation of the references to color in this figure legend, the reader is referred

Coaxial cable-based loopless antenna designs use shortwhips to receive the MR signal, typically ranging from 3 to8 cm [6,20]. The flexible guide wire of the ILMA, on theother hand, can be extended to any suitable lengthdepending on the position of the target tissues. Thisfeature permits very large longitudinal coverage formultislice imaging.

To validate the feasibility and performance of the ILMA,common FSE sequences were used for in vitro and in vivoexperiments. In the results of the phantom test, the presenceof a signal void at the center of the infusion tube likelyresulted from decreased MR signal caused by the flowingvoid effect. However, this effect did not influence thedetection of the vessel wall.

In order to demonstrate the longitudinal coverage of theILMA, an imaging experiment with a cucumber as thesubject was performed by adhering the guide wire to the sideof the cucumber. The guide wire touching air may introducemore noise in the experimental procedure. Thus, the SNR ofthe cucumber was less than that of the phantom.

The reason for the different parameters between thephantom and the cucumber was that the purposes and sizesof the two objects are different. The purpose of the phantomtests was to evaluate the feasibility of the ILMA for vesselwall imaging, whereas that of the cucumber tests was to

O Array Coil (GE) (A and B) was used as a reference of the aorta position byby the ILMA is reliable. The magnified area enclosed by the square in (A) ise blue arrow indicates the vein. The axial image (C) and sagittal image (D) ofible for small vessel wall imaging. The green arrow points to the aorta wall,nts to the abdominal aorta lumen and the blue arrow indicates the vein. (Forto the web version of this article.)

5H. Yuan et al. / Magnetic Resonance Imaging xx (2012) xxx–xxx

validate the large longitudinal coverage of the ILMA.Furthermore, for the in vivo experiment as demonstrated inFig. 5, the axial image based on a commercial 3.0-T eight-channel TORSO Array Coil was used as a reference. Thepurpose of this was to show that the vessel position wasright as obtained by the ILMA. The imaging qualities werenot compared between the eight-channel TORSO ArrayCoil and the ILMA. Therefore, using the same parameterswas not necessary.

In this work, three healthy rabbits were used todemonstrate the feasibility of the proposed unshieldedguide wire-based ILMA for vessel wall imaging. Furtherexperiments utilizing animal plaque models are thereforeneeded to evaluate the imaging quality of pathologicalvessel wall conditions. In order to minimize motion andsusceptibility artifacts, imaging of the artery walls in thisstudy was performed by inserting the guide wire into theveins adjacent to the abdominal arteries. This approach isnot applicable for situations in which the arteries do nothave concomitant veins. In practice, when inserting anantenna directly into an artery, the guide wire may adhereto the artery wall, thus leading to potential MR signal lossin areas of the local artery wall. In previous studies [6,11],an inflated balloon catheter was used to position the guidewire in the center of the target artery. Further efforts aretherefore warranted to both reduce the effects ofsusceptibility [21,22] artifacts originating from the mono-pole antenna probe and integrate the guide wire with aballoon catheter. In our experiment, the optimal probelength was based on empirical tests rather than ontheoretical prediction. For further clinical pathologicalimaging, it will be necessary to maximize SNR bytailoring the guide wire length to the tissue wavelengthoriginating from the varying dielectric permittivity indifferent biological tissues.

Finally, there is a question about safety. A metallic wireplaced on the animal body can cause heating at the tip of thewire during RF transmission [23]. However, if the guidewire is less than 60 cm, the temperature changes are less than7°C [23]. If the antenna is decoupled during RF transmis-sion, the temperature changes are less than 1°C [4]. Inpresent study, we used two means to avoid the heating at thetip of the guide wire. First, we used a length of 58.7 cm (lessthan 60 cm) for the guide wire. Second, we used a PIN diodefor the ILMA decoupling.

In conclusion, we reported a novel design of anintravascular loopless monopole antenna for MRI of thevessel walls. The guide wire of the proposed ILMA isunshielded, thus allowing a much thinner and flexibleguide wire for applications in small vessels. Theexperimental results indicate that the proposed ILMA canbe used to image target vessel walls with a largelongitudinal coverage and high resolution. The proposedILMA may therefore be suitable for MR-based interven-tional diagnoses and therapeutic solutions for cardiovascu-lar atherosclerotic diseases.

Acknowledgment

The authors would like to thank Hua Zhong, Ph.D., andBrad Manor, Ph.D., for their editorial assistance, Huidong Guat GE China Healthcare for his technical assistance, BenshengQiu at Washington University for his valuable suggestionsregarding this work and Yucheng Zhang at Beijing AnzhenHospital for his advice in guide wire selection.

References

[1] Gersh BJ, Sliwa K, Mayosi BM, Yusuf S. The epidemic ofcardiovascular disease in the developing world: global implications.Eur Heart J 2010;31:642-8.

[2] Gu H, Zhang F, Meng Y, Qiu B, Yang X. Development of a 0.014-in.,anti-solenoid loop MR imaging guidewire for intravascular 3.0-T MRimaging. Magn Reson Imaging 2011;29:1002-6.

[3] Zhang X, McKay CR, Sonka M. Tissue characterization inintravascular ultrasound images. IEEE T Med Imaging 1998;17:889-99.

[4] El-Sharkawy A-MM, Qian D, Bottomley PA. The performance ofinterventional loopless MRI antennae at higher magnetic fieldstrengths. Med Phys 2008;35:1995-2006.

[5] Fayad ZA, Fuster V, Fallon JT, Jayasundera T, Worthley SG, Helft G,et al. Noninvasive in vivo human coronary artery lumen andwall imagingusing black-blood. Magn Reson Imaging Circ 2000;102:506-10.

[6] Qiu B, Karmarkar P, Brushett C, Gao F, Kon R, Kar S, et al.Development of a 0.014-inch magnetic resonance imaging guidewire.Magn Reson Med 2005;53:986-90.

[7] Quick HH, Debatin JF, Ladd ME. MR imaging of the vessel wall.Radiology 2002;12:889-900.

[8] Ocali O, Atalar E. Intravascular magnetic resonance imaging using aloopless catheter antenna. Magn Reson Med 1997;37:112-8.

[9] McKinnon GC, Debatin JF, Leung DA, Wildermuth S, Holtz DJ, vonSchulthess GK. Towards active guidewire visualization in interven-tional magnetic resonance imaging. Magn Reson Mater Phys Biol Med1996;4:13-8.

[10] Kantor HL, Briggs RW, Balaban RS. In vivo 31P nuclear magneticresonance measurements in canine heart using a catheter coil.Circulation 1984;55:261-6.

[11] Zimmermann GG, Erhart P, Schneider J, vonSchulthess GK, SchmidtM, Debatin JF. Intravascular MR imaging of atherosclerotic plaque:ex vivo analysis of human femoral arteries with histologic correlation.Radiology 1997;204:769-74.

[12] Quick HH, Ladd ME, Nanz D, Mikolajczyk KP, Debatin JF. Vascularstents as RF antennas for intravascular MR guidance and imaging.Magn Reson Med 1999;42:738-45.

[13] Zimmermann-Paul GG, Quick HH, Vogt P, von Schulthess GK, KlingD, Debatin JF. High-resolution intravascular magnetic resonanceimaging monitoring of plaque formation in heritable hyperlipidemicrabbits. Circulation 1999;99:1054-61.

[14] Quick HH, Ladd ME, Zimmermann-Paul GG, Erhart P, Hofmann E,von Schulthess GK, et al. Single-loop coil concepts for intravascularmagnetic resonance imaging. Magn Reson Med 1999;41:751-8.

[15] Atalar E, Bottomley PA, Ocali O, Correia LCL, Kelemen MD, LimaJAC, et al. High resolution intravascular MRI and MRS by using acatheter receiver coil. Magn Reson Med 1996;36:596-605.

[16] Susil RC, Yeung CJ, Atalar E. Intravascular extended sensitivity(IVES) MRI antennas. Magn Reson Med 2003;50:383-90.

[17] Yang X, Bolster BD, Kraitchman DL, Atalar E. Intravascular MR-monitored balloon angioplasty: an in vivo feasibility study. J VascInterv Radiol 1998;9:953-9.

[18] Takehiko Tsukiji, Masaaki Yamasaki, Yasunori Kumon. Double-folded monopole antenna with coaxial cable. PIERS Proceedings,Moscow, Russia, August 18–21, 2009; pp. 861–864.

6 H. Yuan et al. / Magnetic Resonance Imaging xx (2012) xxx–xxx

[19] Hamirani YS, Avanes E, Kadakia J, Nasir K, Patel A, BudoffMJ. Association of coronary artery diameter on non contrastcardiac computed tomography with coronary artery calcium scoresand traditional risk factors. J Am Coll Cardiol 2010;55 A71.E666.

[20] Meng Y, Zhang F, Gu H,Wang J, Yuan C, Zhang Z, et al. Intravascular3.0 T MRI using an imaging-guidewire: a feasibility study in swine.Appl Magn Reson 2010;40:105-12.

[21] Andris P, Frollo I. Optimized measurement of magnetic field mapsusing nuclear magnetic resonance (NMR). Meas Sci Technol 2011;22:1-8.

[22] Andris P, Frollo I. Measurement of magnetic field with backgroundusing a low-field NMR scanner. Meas Sci Technol 2012;23:1-6.

[23] Liu C-Y, Farahani K, Lu DSK, Duckwiler G, Oppelt A. Safety ofMRI-guided endovascular guidewire applications. J Magn ResonImaging 2000;12:75-8.